The Long-Term Effect of Slavery on Violent Crime: Evidence from US Counties* Moamen Gouda† Anouk S. Rigterink‡ Hankuk University of Foreign Studies London School of Economics and Political Science Preliminary Draft 16 March 2016 Abstract: This study investigates the long-term relationship between slavery and violence in USA. Although considerable qualitative evidence suggests that slavery has been a key factor behind the prevalence of violence, especially in Southern USA, there has been no large-N study supporting this claim so far. Using county-level data for the USA, we find that the proportion of slaves in the population in 1860 is associated with an increase in the rate of violent crimes in all census years for the period 1970-2000. This relationship is robust to including state fixed effects, controlling for various historical and contemporary factors, as well as to instrumenting for slavery using environmental conditions. We explore two potential channels of transmission: (1) slavery leading to higher levels of inequality, which could increase violent crime, and (2) slavery contributing to an ingrained culture of Southern violence. Our results show that only the proportion of slaves living on large slave holdings, as opposed to small slave holdings, is related to contemporary violent crime, supporting inequality as a channel of transmission. We find some tentative evidence supporting culture of violence between the white and black population as a second channel of transmission. JEL Classification: J15, J71, K42, N31, Z13 Keywords: Slavery, crime, inequality, culture, violence, US South * We would like to thank two anonymous reviewers, Boban Aleksandrovic, Seo-Young Cho, Tom Gobien, Jerg Gutmann, Shima’a Hanafy, Bernd Hayo, Helmut Leipold, Florian Neumeier, Nathan Nunn, John Pfaff, Niklas Potrafke, Elisabeth Schulte, and Christian Traxler for their helpful comments and suggestions. Special thanks to Graziella Bertocchi and Arcangelo Dimico for sharing some data. We also thank the participants at the 2013 MACIE brown bag summer seminar in Marburg for their helpful comments. This work was supported by Hankuk University of Foreign Studies research grant. The usual disclaimer applies. † Graduate School of International and Area Studies, Hankuk University of Foreign Studies. Email: [email protected]. Phone: +82(0)1099752712 ‡ Department of International Development, London School of Economics and Political Science. Email: [email protected]. 1. Introduction Since the 18th century, it is noticeable that violence is more prevalent in Southern USA than in other parts of the United States (Ayers, 1991; Nisbett, 1993). Clarke (1998, p. 275) states that “Violence was as much a part of the Southern landscape and culture as azalea festivals and bourbon whiskey”. A considerable body of literature investigates the reason behind the prevalence of violence in South USA, a phenomenon coined as “Southern violence” (Hackney, 1969; Gastil, 1971; Wasserman, 1977). Southern violence continues up till today. In 2011, the South was the region with the highest violent crime rate (Federal Bureau of Investigation, 2011). According to the 2011 U.S. Peace Index that measures the level of peacefulness, or “absence of violence” at the state level, the South was the least peaceful region in the United States, having nine of the ten nationally most violent states (Institute for Economics and Peace, 2011). Many studies hypothesize that the institution of slavery was an important factor behind this culture of Southern violence (Nash, Jeffrey, Frederick, Davis, & Winkler, 2003, p. 362; Cardyn, 2002; West, 2012). Considerable empirical research investigates the long-term effect of slavery on various economic outcomes. Recent findings indicate that slavery has a persistent and long-lasting effect on income inequality (Bertocchi & Dimico, 2014), economic development (Acemoglu, Johnson, & Robinson, 2002; Nunn, 2008; Maloney & Caicedo, 2016), racial educational inequality (Bertocchi & Dimico, 2012), and political attitudes (Acharya, Blackwell, & Sen, Forthcoming). Investigating the long-run development of different municipalities in Colombia, Acemoglu, García-Jimeno, and Robinson (2012) find that the historical presence of slavery is associated with an increase in poverty rate and a reduction in school enrollment, vaccination coverage, and public good provision. This study contributes to the economic and sociological literature by empirically investigating the long-term relationship between slavery and violent crime in the US, especially in Southern USA. We propose two potential channels of transmission between 19th century slavery and contemporary violent crime; firstly, slavery led to higher levels of inequality, which could increase violent crime. Secondly, slaveholders’ reliance on coercion to control slaves may have contributed to an ingrained culture of violence, which contributed to the prevalence of southern violence. Regarding the first channel of transmission, Engerman and Sokoloff (1997; 2002) argue that the existence of slavery in Southern USA led to significant inequality between different segments of the population. Although slavery was formally abolished in 1865, this inequality persisted over time. Engerman and Sokoloff argue that persistent inequality has negative consequences in terms of economic development in the long run (Engerman & Sokoloff, 1997; 2002). Persistent inequality affected other important institutions such as patents (Khan & Sokoloff, 1998), suffrage (Engerman & Sokoloff, 2005), provisions of primary education (Mariscal & Sokoloff, 2000) and taxation (Sokoloff & Zolt, 2007). As there is considerably 2 evidence for a link between inequality and violent crime, (Fajnzylber, Lederman, & Loayza, 2002; Wilkinson, Kawachi, & Kennedy, 1998; Kelly, 2000; Blau & Blau, 1982), we extend Engerman and Sokoloff’s (1997; 2002) argument by hypothesizing that slavery contributes to prevalence of violence in Southern USA through persistent inequality. As for the second transmission channel, Hackney (1969) and Gastil (1971) argue that Southern violence can be attributed primarily to a unique cultural pattern prevalent in the South which persists, despite considerable economic and structural change in this region, to produce a consistently high rate of interpersonal violence. Gastil (1971) declare that the degree of ‘Southernness’ in the culture is a more powerful predictor of violence than socioeconomic factors, such as educations, age, or economic status. A considerable body of research, mainly sociological, hypothesizes that Southern violence stems from specific cultural factors (Bruce, 1979; McCall, Kenneth, & Cohen, 1992; Clarke, 1998; Dixon & Lixotte, 1987; Ellison, 1991; Hayes & Lee, 2005). The practice of slavery, a number of qualitative studies argue, is a key factor behind this culture of violence (Cash, 1941; Franklin, 1956; Gastil, 1971; Wyatt-Brown, 1986). This leads us to hypothesize that culture is a channel of transmission between historical slavery and present-day violence. Using county-level data for the USA, we find that the proportion of slaves in the population in 1860 is positively related to violent crime in all census years from 1970 to 2000. This relationship is robust to including state fixed effects, controlling for a number of historical and contemporary factors, including historical income and inequality, and contemporary unemployment, income inequality, distance to the Mexican border and proportion of youth in the population, as well as to using instrumenting for slavery using environmental conditions. Exploring the two potential channels of transmission, we find that the proportion of slaves in the population residing on large slave holdings is related to contemporary violent crime, but the proportion of slaves on small slave holdings is not, suggesting inequality as a channel of transmission. However, our results show some tentative evidence supporting culture of violence between the white and black population as a second channel of transmission. This study is divided into six sections. The next section provides a theoretical background on the relation between the legacy of slavery and violence and presents our hypotheses. Section 3 presents the estimation strategy and data. Section 4 presents results on the empirical relationship between slavery and contemporary violent crime. Section 5 presents results on the two hypothesized channels of transmission between slavery and violent crime, namely inequality and culture of violence. Section 6 concludes. 3 2. Theoretical Background and Hypotheses There has been a recent interest in investigating the long-term effect of historical institutions on violent behavior. Jha (2008) argues that inter-ethnic medieval trade has left a lasting legacy on the patterns of religious violence between Hindus and Muslims in India. Voigtländer and Voth (2012) demonstrate that the same places in Germany that saw violent attacks on Jews during the plague also showed more anti-Semitic attitudes over half a millennium later. As for the United States, a considerable body of sociological literature investigates the reason behind the prevalence of violence in the South, a phenomenon coined as “Southern violence” (Hackney, 1969; Gastil, 1971; Wasserman, 1977). Since the 18th century, it was noticeable that the violence was far more prevalent in the South than in other parts of the United States (Ayers, 1991; Nisbett, 1993). Messner, Baller and Zevengergen (2005, p. 633) state that “distinctive